Titanic survival predictions

Level 1:

1. What do the following two pictures illustrate?

A、
The survival rate of children under 10 years old is relatively high, and it has nothing to do with the cabin level.
B、
First class passengers aged 20-50 have a higher survival rate
C、
For men, the older they get, the lower their survival rates
D、
Women have higher survival rates

Answer: BCD

2、

What does this picture illustrate?

A、
The survival rate of passengers disembarking at Port C is higher than 55%
B、
Passengers who boarded the ship at Port S had the lowest survival rate
C、
Port C has the largest number of people boarding the ship.
D、
Port S is the port with the fewest number of people boarding the ship.

Answer: AB

3、

What do the following four pictures illustrate?

A、
Although many first-class rich people basically boarded the ship at Port S, Port S had the lowest survival rate. This may be because many of the people who boarded the ship at Port S were third-class passengers.
B、
Most of the people boarding the ship at Port S are second-class passengers.
C、
More than 90% of the people boarding the ship at Port Q are third-class passengers
D、
The survival rate is highest when boarding the ship at Port C

Answer: ACD

4、

What do the following two pictures illustrate?

A、
The survival rate for people traveling alone on a boat is about 34%
B、
If a sibling or loved one is also on board, the survival rate is the highest.
C、
When the number of siblings or lovers is relatively large, the survival rate tends to decrease.

Answer: ABC

Level 2: Fill in missing values

import pandas as pd
import numpy as np

def process_nan_value(data):
    '''
    处理data中缺失值，有缺失值的特征为`Age`，`Cabin`，`Embarked`。
    :param data: 训练集的特征，类型为DataFrame
    :return:处理好缺失值后的训练集特征，类型为DataFrame
    '''

    #********* Begin *********#
    data['Age'].replace(np.nan,np.nanmedian(data['Age']),
       inplace=True)
    #data['Age'].replace(np.nan,np.nanmedian(data['Age']),inplace=True)
    data.drop(labels='Cabin',axis=1,inplace=True)
    data['Embarked'].replace(np.nan,'S',inplace=True)

    return data
    #********* End *********#

Level 3: Feature Engineering and Survival Prediction

import pandas as pd
import numpy as np
import sklearn
from sklearn.ensemble import RandomForestClassifier
from sklearn.ensemble import RandomForestRegressor

#********* Begin *********#
titanic = pd.read_csv('./train.csv')
def set_missing_ages(df):

    # 把已有的数值型特征取出来丢进Random Forest Regressor中

    age_df = df[['Age', 'Fare', 'Parch', 'SibSp', 'Pclass']]

    # 乘客分成已知年龄和未知年龄两部分

    known_age = age_df[age_df.Age.notnull()].values

    unknown_age = age_df[age_df.Age.isnull()].values

    # y即目标年龄

    y = known_age[:, 0]

    # X即特征属性值

    X = known_age[:, 1:]

    # fit到RandomForestRegressor之中

    rfr = RandomForestRegressor(random_state=0, n_estimators=2000, n_jobs=-1)

    rfr.fit(X, y)

    # 用得到的模型进行未知年龄结果预测

    predictedAges = rfr.predict(unknown_age[:, 1::])

    # 用得到的预测结果填补原缺失数据

    df.loc[(df.Age.isnull()), 'Age'] = predictedAges

    return df

titanic = set_missing_ages(titanic)

dummies_Embarked = pd.get_dummies(titanic['Embarked'], prefix= 'Embarked')

dummies_Sex = pd.get_dummies(titanic['Sex'], prefix= 'Sex')

dummies_Pclass = pd.get_dummies(titanic['Pclass'], prefix= 'Pclass')

df = pd.concat([titanic, dummies_Embarked, dummies_Sex, dummies_Pclass], axis=1)

df.drop(['Pclass', 'Name', 'Sex', 'Ticket', 'Cabin', 'Embarked'], axis=1, inplace=True)

# print(df)

train_label = df['Survived']

train_titanic = df.drop('Survived', 1)

titanic_test = pd.read_csv('./test.csv')
titanic_test = set_missing_ages(titanic_test)

dummies_Embarked = pd.get_dummies(titanic_test['Embarked'], prefix= 'Embarked')

dummies_Sex = pd.get_dummies(titanic_test['Sex'], prefix= 'Sex')

dummies_Pclass = pd.get_dummies(titanic_test['Pclass'], prefix= 'Pclass')

df_test = pd.concat([titanic_test,dummies_Embarked, dummies_Sex, dummies_Pclass], axis=1)

df_test.drop(['Pclass', 'Name', 'Sex', 'Ticket', 'Cabin', 'Embarked'], axis=1, inplace=True)
#model = RandomForestClassifier(random_state=1, n_estimators=10, min_samples_split=2, min_samples_leaf=1)
model = RandomForestClassifier(n_estimators=10)
model.fit(train_titanic, train_label)

predictions = model.predict(df_test)

result = pd.DataFrame({'Survived':predictions.astype(np.int32)})

result.to_csv("./predict.csv", index=False)


#********* End *********#

Titanic survival predictions

Level 1:

Level 2: Fill in missing values

Level 3: Feature Engineering and Survival Prediction

Guess you like